ARM의 CEO가 팟캐스팅, AI 기회 및 업계 과제에 대해 이야기합니다.

### AI 및 칩 제조 분야에서 ARM의 전략적 발전과 기회
* 🎙 ARM의 CEO는 최근 업계 리더들과 소통하고 기술 환경에 대한 인사이트를 공유하기 위해 **Tech Unheard라는 팟캐스트**를 시작했으며, 첫 번째 게스트로는 **Jensen Huang**이 참여했습니다.
* 💼 ARM은 소프트뱅크에 약 **320억 달러에 인수**되었으며, 1년여 전 성공적인 기업공개 이후 현재 시장 가치는 **약 1,600억 달러**에 달합니다.
* 🔄 ARM의 비즈니스 모델은 **수평적 라이선스 접근 방식에서 데이터 센터 및 자동차 같은 분야에 특화된 제품에 초점을 맞춘 수직적 전략**으로 진화하여 새로운 시장을 공략할 수 있게 되었습니다.
* 🤖 전 세계 인구의 70%가 ARM 기술과 상호 작용하는 가운데, **모든 AI 워크로드가 궁극적으로 ARM 아키텍처에서 실행**될 것이기 때문에 AI가 엄청난 성장 기회를 제공한다고 믿습니다.
* 📈 엔비디아와의 경쟁에도 불구하고 **ARM은 이미 고급 데이터 센터에 통합**되어 있으며, CEO는 **AI에서 ARM의 역할이 다양한 디바이스에서 알고리즘 학습에서 추론**으로 확장될 것으로 예상하고 있습니다.
* 🌍 CEO는 특히 **TSMC의 지배력으로 인한 칩 제조 집중에 대한 우려**를 표명하며 보다 다양하고 지리적으로 분산된 반도체 산업을 옹호했습니다.
* 💡 그는 특히 선도 기업의 경우 **연간 최대 350억 달러에 달하는 새로운 팹에 필요한 막대한 자본 지출**을 고려할 때 칩 제조 투자에 대한 정부의 참여가 중요하다고 강조했습니다.
* ⚖ ARM은 퀄컴과의 법적 문제를 해결해 나가는 과정에서 업계 내 경쟁의 복잡성을 강조하며, **ARM은 시장 수요를 효과적으로 충족하는 솔루션**을 제공하는 데 집중하고 있다고 말했습니다.

AI가 기술의 모든 측면에 통합됨에 따라 ARM은 엄청난 성장을 이룰 준비가 되어 있습니다.

That's why I have you. And thank you all for joining us. We have a lot of questions for you, so we'll go ahead and get started. First, I begin. You are not just the CEO of aam; you're now a podcaster. All right, you began in your spare time, I guess. You began a new podcast called Tech Unheard, and your first interview was with Jensen Huang. That's right, which I gave a listen to. First of all, what gave you the idea to start a podcast? And secondly, how do you find the time as CEO to do something like that too? Yeah. So first off, thank you again for having me here. I'm an amateur at this, so you don't have to worry about your day job as far as podcasting goes. You know, we were talking with some of our board members about how to use a unique medium to talk to leaders, talk about the armed story. And one of the ideas that was suggested

was a podcast. I was a radio disc jockey, actually spinning records in university. So, I had a little bit of experience and aspirations to do that kind of work early in my life. So, in talking with some folks, you know, around the team, we said, you know, let's try this and let's do something a little bit more innovative and talk to other leaders and give them a perspective of maybe a conversation that they may not naturally have with someone who's professional like you are. So, yeah, we kicked it off. Please listen, it's on Apple Music, Spotify, and Jensen was my first guest. Great, great. Congratulations. And, you know, in terms of the business arm, was acquired by SoftBank for about $32 billion. At some point, SoftBank agreed to sell it for about $40 billion. That deal didn't go through, of course, in video was supposed to buy it.

That deal didn't go through. Now, you went public just over a year ago, companies worth roughly $160 billion, which seems like a successful IPO. But could you talk to us about what the business model for ARM has been and how it's evolved over time, over that period of time? And especially, what's the strategy going forward? Yeah, well, thank you for the you know, for the nice words. I'm a 30-year-old plus company. Our business model for most of our time on earth, if you will, was very horizontal in that we had a licensing model. We develop IP. Our products are primarily CPUs. The business model is very horizontal in nature, meaning that we designed a very general-purpose product, very power-efficient product, and then license it to customers independent of the vertical market. And in other words, we didn't really target a certain vertical. I took over the core IP business right after SoftBank bought ARM, and we pivoted

to a more of a vertical approach, meaning that we now have products very specific for the data center, very fitting for automotive, very big variety. Smartphones were born. And that diversification really allowed us to expand into many new markets with brand new products. SoftBank buying ARM in 2016 allowed us to invest in those new areas. So fast forward the result. What you're seeing now in terms of performance is that strategy. You know, going forward as we think about the next number of years, what we're seeing in the marketplace is a need for more solutions. Chips are getting increasingly more difficult to build. It takes a long time to design a chip, takes even longer to fabricate the chip. So the more that arm can do to help customers get to market faster is helpful. So we're now developing solutions, what we call successes that get people to market sooner and allow them to get more profitably in a sooner way.

Mm hmm. Okay. I want to talk about A.I. and, in particular, the opportunities in AI. There is some debate, as we've referred to, I think, in a couple of these sessions, about whether there's a bubble. And I'm going to assume that you think that there are real opportunities there. But can you talk about what ARM's role is there and how you see the company evolving to take advantage of those opportunities? So, one of the most unique things about our company is the fact that the device that we build, the CPU. We estimate that 70% of the world's population touches ARM in some way. We're in security cameras, automobiles, PCs, smartphones, data centers. And what that means is every workload, whether it's an application, an operating system, or A.I., runs through ARM in some way. So, that means A.I. is going to run on ARM, period. And we're the only company on the planet

that can run those A.I. workloads from the smallest devices on the edge. Again, a wearable glasses watch phone to the data center. So for us, A.I. is a gigantic opportunity in terms of growth going forward because AI is going to be everywhere. And as far as a bubble goes, I just don't ascribe to that. I look at things such as when people say, Gosh, is it overhyped? Or the stocks have gone down last quarter. That's almost like saying, Should I short AT&T in 2000 because the Internet is not going to happen? Or if Ford was a stock in 1907, would I short them? Because the automobile is not going to happen. It's just not. I mean, I have seen such advancement across the air in the last couple of years that the amount of innovation that it's going to drive is just going to be incredible. Okay. Okay.

The chip company that probably has taken the most opportunity has been able to capitalize the most on AI so far as in video chip maker in your space. They now have a near monopoly in terms of the highest-end AI accelerators. How do you see that evolving in the future? And are there opportunities for ARM in particular to be able to play in that space and drive real revenue from? Yeah. So they're already invidious most advanced chip called the GB 200, which is Grace Blackwell. Microsoft just announced that they're going to be deploying Grace Blackwell, the most advanced Nvidia chip in their data centers that uses the ARM CPU. So Grace is the ARM CPU. BLACKWELL is the Nvidia GPU. So ARM is there already. So we are going to play in the most advanced data centers using an AI ARM. ARM will be there. I think what's most exciting for us, as I mentioned though, is the fact that not only will we be in the data center for

training these algorithms. Ultimately, the training has to be transferred into inference, and training is the teacher. Inference is the student. Inference is actually running the workload and inference actually running that AI agent that's going to happen in the data center. That's going to happen in the car. That's going to happen on your watch. That's going to happen in glasses. That's going to happen in phones and everywhere it happens. It's going to run on arm. So it's going to be a gigantic opportunity for us. The data center is just one place. Mm hmm. Okay. So you see arm having opportunities kind of from the beginning to the end, more than often, more than opportunities. The agents will run locally. And the reason for that is you'll have a hybrid situation where some of that will run in the cloud. Some of it will run in your local device. Again, your glasses, your wearable, your phone.

But locally, what you'll be able to add is security, things that can make sure it's private to you, that not all your data is being passed to the cloud and vice versa. So yeah, for ALM, I think it's going to be just an amazingly large opportunity. Okay. As we were talking about before, I just moved from Japan, where we spent a lot of time looking at SoftBank and especially Masayoshi Son's ambitions. We've written stories about how SoftBank is looking at ways to enter the chip market and particularly to compete within video quite directly in GPUs. He has a secret plan, his analogy that we've written about in the past. How does ARM, isn't still 90% owned by SoftBank? How does ARM fit into that equation? Yeah, so no product launch announcements today. So nothing I can say about products coming out that you've written about. I can say that on the SoftBank board member, I do advisory work for SoftBank.

I talked to Masa all the time. It's no secret that he is a big believer in AI, has been for quite some time. He's talked about it, actually, probably as long as any of the folks out there. I think his vision, in terms of where it's all coming together now, you can start to see the pieces of it, as I just described. Where does aam fit? All those A.I. workloads are going to run on aam somewhere, somehow. So, that's the reason that we spend a lot of time talking to SoftBank about the future. Okay. Okay. Earlier, we had the ASML CEO on, and he was talking about some of the challenges that they're facing in the market. They surprised investors last week, as we talked about. And it does seem that there's a little bit of a disconnect in the market, where many people see opportunities. ASML reported that their orders were about half what analysts had thought they were going to be.

Seems there are certain areas of the chip industry that are struggling quite a bit. And on the other end, you have Sam Altman talking about how we need more capacity in the chip industry. We need billions of dollars, maybe trillions of dollars of investment. What is the disconnect, and how does that get resolved? Yeah, so there's a lot to unpack there. Where ARM plays, as I mentioned, because 70% of the world's population uses ARM, and I think 300 billion chips have shipped since we started that have ARM inside. And every year, about 30 billion chips. We have a unique view into the industry because just about every digital device uses ARM. There are some markets that aren't accelerating as fast as others, and candidly, those are the ones that aren't using A.I. in a very large way. But in areas that can use A.I. and will use A.I., people can't go fast enough.

And I'll give you an example. Many of the hardware devices that are being introduced today were designed two or three years ago. That means the chips were designed two or three years ago. The chips that are going in your phones or your wearables. Those were all invented before CBT was released or before Lamo was released. So what we're seeing on one end of the industry is just a rush to develop more and more chips that have this air capability. Now it takes time. You know, chips aren't built overnight. In fact, it takes from the time we engage a partner, you know, two or three years at least until that comes to market. So I think you'll see some of these new products coming out. I think what Sam is talking about is just generally the ceiling to get to AGI. Artificial general intelligence requires more compute. And the more compute you have, the better the models get. So I see. I see both.

But generally speaking, I'm pretty optimistic about the growth for our industry. Okay. Okay. We were talking about this a bit before. One of the gating factors is making sure that you have chip manufacturers who can make the most advanced chips. TSMC seems to be gaining market share right now as a couple of the key competitors struggle, Intel in particular. And Samsung also has been struggling a bit. They're both trying to move into the foundry business a bit in competition with TSMC. What are the dangers around having TSMC garner so much market share in foundry? And is it good for the industry, or would it be better if there's a more diversified manufacturing base for the industry? Yeah, TSMC is an amazing company. They've just done so much for the industry relative to the innovation, how they work with partners, etc. etc. If you remove semiconductor manufacturing or foundry for a moment

and just step back and say, is it healthy for any industry to have one dominant supplier and then add on top of that a dominant supplier that's largely concentrated in one part of the world? Not, not really. So as a result, I think it's necessary just in terms of diversity of supply base, to have more and more advanced manufacturing and geographic diversity. Fabs. The United States. Fabs in Europe. Fabs in India. Fabs in Saudi Arabia. I think over time, we're going to see not only more suppliers, but geographic diversity as well. Okay. Governments obviously are investing a lot in that very question. You're seeing the United States do it. You're seeing Europe do it. Japan has spent a lot of money on it, too. Do you think that's making progress at this point? Do you see light at the end of that tunnel? It's the capital expenditures

required for these fabs are enormous. TSMC, Intel, Samsung, I think their CapEx is anywhere between 30 to $35 billion a year in new factories. And if you start looking at the revenues that these companies produce, you start very quickly looking at, oh my gosh, that is a lot of money to invest. Now, on one hand, you can look at it and say, well, the demand will be there, etc., etc. But it doesn't take much in terms of a year slowdown in demand and the companies have a really negative financial outcome. So, yes, governments have been involved, I think should be involved in my opinion. We saw in the U.S. for the very first time a policy act around CHIPS Act to put funding into the fabs, which I generally a big supporter of. I think it's necessary. Okay. Let's talk a little bit more about Intel and their struggles. You grew up in a chip industry where

Intel was really the dominant player, kind of set the pace for technology evolution over time. Now we're openly talking about whether Intel is going to be broken up, maybe sold into pieces. What kind of opportunities are there for ARM within that space as Intel goes through these struggles? Yes, we work very closely with Intel. It's a bit of strange bedfellows. Many folks would consider ARM and Intel to be natural competitors on the product space. But what's very important to us, back to this discussion around supply base, is that Intel Foundry Services (IFRS) is successful. We do a lot of work with Intel as we do with Samsung and TSMC to ensure that our products can be manufactured there. We work very closely with their engineers to ensure that the leading-edge technology can run on ARM. So we want to see Intel successful, and we want to see if it's successful. I think it might be strange to hear the ARM CEO say that, but we want Intel

manufacturing to be successful. Okay. Bloomberg reported that ARM took a pretty close look at buying a part of Intel. Would you like to comment? Nothing I can say on that one. Okay. Okay. All right. Thank you. Thank you. I wanted to turn. You talked about moving up the value stack, kind of moving beyond its traditional place in the industry. And as you have moved up the value chain, you've gotten closer to doing what some of your customers do. Right? Getting closer to being able to do these complete designs. And. Yes. It speeds up innovation, as you were alluding to, but it also puts you in closer competition with some of your customers. You have a lawsuit with Qualcomm in particular? Yeah. Isn't it in both companies' interest to settle that and to resolve the legal dispute at this point? Or what would you say? These are tough questions. It's a good thing I'm a podcaster, so I know what's coming.

Right. You know, first thing on, competing with our customers, you know, it's rather complicated because if you look at some of our customers, our customers are Amazon. Our customers are Microsoft. Our customers are Apple. Our customers are Tesla. They all build chips using ARM. I'm not going to build an electric car. I'm not going to build a phone. I'm not going to build a data center. So to look at the value chain relative to who builds chips, relative to whether your end business is a chip business or a product business. It's gotten a lot more gray. We follow what the industry is demanding, and what the industry wants to see is solutions getting to market faster. And that's what we're focused on. Qualcomm. Not much I can say on that, other than we're headed to a trial. I think it's the third week in December. We feel very good about our case.

We think our case is quite simple and straightforward. And as I said, December, we'll find out. Okay. So you head to the courtroom then. One last question. I think we have time for this. In this audience in particular, there's a lot of interest in whether ARM, which is historically a British company, whether they would ever look at doing some sort of dual listing. Of course, we've heard about this many times in the past. A year ago when you did your IPO, you decided you would do it in the U.S. instead. One of the opportunities for listing here, what are the pluses and minuses of that sort of thing? Yeah, we did. We listed in New York over a year ago. And at that time our comments were we would look at a secondary listing. We're still open to it. It's not something that's top of mind right now, quite candidly.

But we'll continue to have discussions with both stakeholders in the government and in the local exchange. We'd love to find a way at some point in time, but also at the same time, there's a lot of other things that keep me busy relative to my day job. Okay. All right. Rene Haas, thank you very much. Please join me in welcoming. Thank you. Thank you.

Arm CEO on Intel, Chips, AI, Listing in US

Welcome to the Bloomberg Technology YouTube channel. Subscribe for original series and mini-docs on game-changing technologies, profiles on innovators, live conversations with experts, and breaking technology news & analysis from Bloomberg TV.

Bloomberg Technology

CEO의 여정: ARM에서 팟캐스팅까지

ARM의 CEO는 라디오 디스크 자키였던 자신의 과거 경험을 바탕으로 리더들과 독특한 대화를 나누는 팟캐스트 'Tech Unheard'를 시작했습니다.

ARM의 비즈니스 모델은 수평적 라이선스 접근 방식에서 데이터 센터 및 자동차 같은 특정 시장을 대상으로 하는 수직적 전략으로 발전했습니다.

이제 ARM은 고객이 더 빠르고 수익성 있게 칩을 제작할 수 있도록 지원하는 솔루션을 제공하는 데 주력하고 있습니다.

ARM의 CPU는 전 세계 디바이스의 70%에 탑재되어 있으며, 엣지 디바이스에서 데이터 센터에 이르기까지 AI 워크로드를 실행하는 데 필수적인 역할을 하고 있습니다.

ARM은 이러한 독보적인 위치를 바탕으로 다양한 디바이스에서 학습 및 추론의 성장 기회와 함께 증가하는 AI 수요를 활용할 수 있습니다.

ARM은 이미 파트너십을 통해 고급 데이터센터에 참여하고 있으며, 클라우드와 로컬 디바이스 간의 AI 워크로드 분산 하이브리드 모델은 상당한 수익 잠재력을 제공합니다.

소프트뱅크가 90% 지분을 소유한 ARM은 창립 이래 3,000억 개의 칩을 출하하며 칩 산업에서 중요한 역할을 담당하고 있습니다.

전 세계 인구의 70%가 ARM을 사용한다는 점에서 ARM의 독보적인 위치는 업계의 역학 관계에 대한 통찰력을 제공합니다.

일부 시장은 어려움을 겪고 있지만, AI를 활용하는 시장은 빠른 성장과 칩 용량에 대한 수요 증가를 경험하고 있습니다.

오늘 소개하는 하드웨어 디바이스는 2~3년 전에 설계되었으며, 휴대폰과 웨어러블에 사용되는 칩의 개발 시기도 비슷합니다.

TSMC가 칩 제조 분야에서 시장 점유율을 높여가면서 지배적인 공급업체가 한 지역에 집중될 경우의 위험에 대한 우려가 커지고 있습니다.

각국 정부가 미국, 유럽, 인도, 사우디아라비아에 걸쳐 칩 생산에 투자하면서 첨단 제조업의 지리적 다양성을 추구하는 움직임이 있습니다.

TSMC, 인텔, 삼성과 같은 선도적인 칩 회사는 매년 300억~350억 달러를 신규 공장에 투자하고 있습니다.

칩에 대한 수요는 매우 중요하며, 수요가 둔화되면 이러한 기업들은 재무적으로 부정적인 결과를 초래할 수 있습니다.

ARM은 성공적인 제조를 위해 인텔, 삼성, TSMC와 긴밀히 협력하고 있으며, 가치 사슬을 끌어올리기 위해 퀄컴과 같은 고객과도 경쟁하고 있습니다.

ARM의 복잡한 고객 환경과 향후 상장 계획

ARM의 고객사에는 Amazon, Microsoft, Apple과 같은 주요 기업이 포함되어 있어 비즈니스 모델이 복잡합니다.

이 회사는 업계의 요구에 부응하여 솔루션을 더 빠르게 시장에 출시하는 데 주력하고 있습니다.

ARM은 영국에서 2차 상장에 대한 가능성을 열어두고 있지만, CEO가 다른 비즈니스 문제에 집중하고 있기 때문에 현재로서는 최우선 순위는 아닙니다.